The Swedish COVID‐19 control measures and the national commission report

Sweden's COVID‐19 strategy has received major international attention as it was considered largely unique compared to that of other comparable countries and it has been severely criticised both nationally and internationally. ¹ , ² This criticism, mainly from non‐experts in infectious disease control, received strong media attention and was largely endorsed by leading Swedish newspapers. In contrast, the public mainly supported the Swedish model. The government appointed a commission to hold an in‐depth inquiry. Their reports are commendably summarised and commented upon by Professor Ludvigsson in this issue of Acta Paediatrica. ³ This provides a valuable contribution to the debate. The final report by the Commission largely reiterated previous views and criticism and concluded that Sweden was ill prepared for the pandemic and the response was too slow and weak. This was particularly true during the first COVID‐19 wave in spring 2020, which was the major focus of the report. ⁴

One major aim of the Swedish strategy was to protect the older people and others at high risk of severe, even fatal, disease. The Commission argued that this largely failed, in the nursing homes and during social care at home. These deficiencies were indeed generally recognised early on but proved to be difficult to counteract. Hence despite gradual improvements in protective measures in especially the nursing homes, there was a recurrent high mortality rate during the second wave. The high mortality among the elderly was, however, not unique to Sweden. The mean age of COVID deaths of about 83 years ⁵ was for example similar in neighbouring Nordic countries. The commission argued that higher community transmission was the main cause of the higher mortality in Sweden. ⁴ International data, however, suggest that both, deficiencies in the nursing homes and high transmission rates affected the high mortality rate in Sweden. ⁴

Another significant feature of the Sweden's approach was the focus on voluntary measures and personal responsibility rather than mandatory measures and stricter community lockdown. The Public Health Agency (PHA) did not favour the ‘precaution principle’, which is to introduce the highest rigorous restrictions possible. Referring to the Communicable Diseases Act principle of ‘proportionate and preferably evidence‐based restrictions’ the PHA favoured measures that were more easily sustainable and acceptable to the population. The Commission endorsed the voluntary approach while maintaining that the restrictions should have been more rigorous with a higher degree of lockdown. The Commission also endorsed the internationally partly unique decision to keep schools open for all children up to 16 years of age. This may be perceived as contradictory ³ and, indeed, the general conclusions by the Commission appear more negative than the detailed conclusions. ⁴ The Commission highlights that several restrictions failed: (a) the lack of quarantine measures for all incoming travellers from areas reporting COVID‐19 cases during and after the national sports holidays in late February 2020 (week 9);the feasibility and socio‐economic implications of such rather massive measure are, however, not discussed; (b) the lack of temporary banned travel to Sweden mid‐March; the transmission within Sweden was, however, high already at that time, with only a minor part of the cases being imported and rather negligible 2 weeks later; the expected impact of such a ban is thus highly questionable; (c) the lack of stricter rules regarding gatherings, social events, restaurants, etc., including more use of face masks; the degree of operational impact of such increased restrictions is, however, still debated; d) the lack of greater preventive measures ahead of the second wave; however, according to Our World in Data ⁶ the restrictions were greater in Sweden than in all three neighbouring Nordic countries throughout the second wave, although the highest restriction level was reached only by mid‐November 2020. The commission further argued that the Swedish response was not only too weak but also too late, compared to neighbouring countries. Data on preventive measures, are estimated and summarised by Our World in Data as a composite measure (‘COVID‐19 Containment and Health Index’) including thirteen metrics, from wearing of face masks to vaccination. ⁶ These data, only partly referred to in the report, indicate that the early response to the pandemic was actually quite similar in the four neighbouring Nordic countries. The response reached an estimated 25% index (total lockdown/maximum response = 100%) first in Denmark (11 March) and 1 day later (12 March) in the other three countries. A couple of weeks later the maximum indices were 55% in Sweden and 57%–60% in the other three countries indicating slightly less restrictions in Sweden (e.g. open schools). Importantly, however, there is little evidence that higher degrees of lockdown resulted in faster and more significant reductions of COVID‐19 mortality in European countries. ⁶ , ⁷ In contrast there was a high correlation of cumulative mortality and early COVID‐19 mortality reproductive rates before the impact of restrictions. The reproductive rate was significantly higher in Sweden as compared to especially Norway and Finland and could, therefore, largely explain the differences in mortality rates. An earlier and higher number of imported cases may possibly also have partly influenced the epidemic in Sweden. An additional indication that the measures were potentially sufficient in Sweden is that the mortality rates started to decrease (reproductive rate <1) in all four Nordic countries about 4–5 weeks after restrictions were initiated, albeit more slowly during May in Sweden. More strict measures might have had some impact then. Interestingly, whereas it is reported that Sweden successively copied other countries with more strict restrictions (e g use of face masks during public transport), the overall picture suggests that other countries reduced their degree of lockdown towards a more sustainable approach comparable to the Swedish model (e g open schools) during the later COVID‐19 waves. ⁶ The important epidemiological aspects above, e g differential basic reproductive rates, are barely mentioned or discussed in the report. ⁴

The commission strongly highlighted an unacceptable delay in testing and tracing in Sweden as compared to the other Nordic countries. Denmark was indeed very successful in speeding up the testing, but Norway and Finland were not that different from Sweden. ⁴ , ⁶ Although testing is important in COVID‐19 control especially as screening, the possible case isolation and associated tracing has its main merit early and late in an epidemic. Since the community transmission was already quite high in Sweden by mid‐March, increased tracing would have had limited impact. An additional issue is also the operational drawback with delayed results (often >48 h) when PCR is used as opposed to immediate point‐of‐care test results.

The Commission also concluded that the Government should have provided a stronger lead, been less dependent on the PHA and consulted those with more critical views of the Swedish model. But it is unclear what difference that would have made since these critical views received a lot of media attention and since most expertise and experience in infectious diseases control lay within the PHA and the regional infectious disease centres all interacting to combat t COVID‐19. The governments in Denmark and Norway took a stronger lead, but decisions such as mink culling and erratic border closing with major socio‐economic costs, were made without any evidence of its impact on transmission.

There was only a small mention of the major public health issue of the long‐COVID in the report and how that might have been a rational for specific measures. The COVID‐19 mortality and to some degree health care burden seems to have been the main outcomes of interest for the commission. This also reflected the continuous reporting by the PHA during the epidemic. The socio‐economic costs of a higher degree of lockdown are not discussed in this Editorial, but obviously they had to be considered when deciding on different political strategies for combating COVID‐19. The PHA, however, claimed that their perspective was only the health of the people.

Finally, an added value of the report would have been to include more data and reflections on COVID‐19 and its control measures beyond the first and partly second waves. The report was published towards the end of the fourth wave in late February 2022. During 2021 and after the vaccines were introduced, Sweden was among countries least affected by COVID‐19. Data for the whole period 2020‐2021 suggests a relatively low excess all‐cause mortality (3%–5%) as compared to many other European countries. ⁶ , ⁸ , ⁹ , ¹⁰ Maybe the possibly more sustainable COVID‐19 control approach in Sweden may have influenced this outcome. When compared to the reported COVID‐19 deaths (about 15 000), the excess mortality estimate suggests that almost half of the reported deaths were not caused by COVID‐19 or would have occurred anyway before the end of 2021, from an underlying disease. ⁵ , ⁶

An inquiry into the COVID response and impact was important. The report focused mainly on management issues, policies and decision making and provided detailed descriptive data of the measures that were or were not implemented. There were little causal analyses of how the Swedish model did or did not affect the COVID‐19 related morbidity and mortality in the population. The conclusions reached by the Commission appear to partly portray preconceived critical views and may be questioned. Several issues remain unanswered. My general conclusion is that overall the Swedish model was reasonable. It deserves a better Commission inquiry, which unfortunately includes embarrassing and critical weaknesses.

CONFLICT OF INTEREST

None

Biography

Anders Björkman